\begin{table}[htbp]\centering
\def\sym#1{\ifmmode^{#1}\else\(^{#1}\)\fi}
\caption{Job satisfaction regression results}
\begin{tabular}{l*{6}{c}}
\toprule
            &\multicolumn{2}{c}{US (GSS)}               &\multicolumn{2}{c}{UK}                     &\multicolumn{2}{c}{Germany}                \\
            & No controls         &    Controls         & No controls         &    Controls         & No controls         &    Controls         \\
            &                     &                     &                     &                     &                     &                     \\
\midrule
Unemployment rate*100&      -.0026         &      -.0024         &      -.0047         &      -.0051         &      -.0047\sym{**} &      -.0047\sym{**} \\
            &     (.0033)         &     (.0031)         &     (.0025)         &     (.0026)         &     (.0015)         &     (.0015)         \\
Linear time trend&     -.00009         &     .000056         &      -.0012\sym{*}  &      -.0012         &      -.0014\sym{***}&     -.00086\sym{*}  \\
            &    (.00034)         &    (.00032)         &    (.00055)         &    (.00057)         &    (.00039)         &    (.00038)         \\
\midrule
\(N\)       &          22         &          22         &          28         &          28         &          34         &          34         \\
\bottomrule
\multicolumn{7}{l}{\footnotesize This table displays the results of regressions of the average marginal effects of the year dummies (with and without controls)}\\
\multicolumn{7}{l}{\footnotesize on the unemployment rate and a linear time trend using the GSS, HRS, BHPS/UKHLS, and SOEP data.}\\
\end{tabular}
\end{table}
